Additional file 1 of HIVprotI: an integrated web based platform for prediction and design of HIV proteins inhibitors

Additional file 1. Supporting information including Table S1. Performance of QSAR predictive models on three times randomly picked ~ 10% independent/validation data. These models were developed using remaining ~ 90% data during training/testing respectively for each of the six datasets; Table S2. Details of statistical parameters used for the development of IC50 based QSAR models; Table S3. Details of statistical parameters used for the development of percent inhibition based QSAR models; Table S4. Details of chemical descriptors used in the development of IC50 based QSAR models; Table S5. Details of chemical descriptors used in the development of percent inhibition based QSAR models; Table S6. Details of slopes k (predicted vs. observed inhibition) and k’ (observed vs. predicted inhibition) of the regression lines for the QSAR models; Table S7. Details of Y-randomization test performed on the QSAR models; Figure S1. Chemical space analysis of QSAR studies (Table 5) for Protease (PR) (a, b), Reverse Transcriptase (RT) (c, d) and Integrase (IN) (e, f) respectively; Figure S2. Chemical space mapping outline of (a) Integrase (IN), (b) Protease (PR) and (c) Reverse Transcriptase (RT) inhibitors (percentage inhibition) with internal circle showing clustering and 3-D embedding of compounds, middle circle with exact (zoomed) superimposed cluster and outermost circle with specific MCS of each cluster; Figure S3. Scatter plot depicting the applicability domain for IC50 datasets of (a) Integrase (IN), (b) Protease (PR) and (c) Reverse Transcriptase (RT); Figure S4. Scatter plot depicting the applicability domain for percentage inhibition datasets of (a) Integrase (IN), (b) Protease (PR) and (c) Reverse Transcriptase (RT).